Military Model III

[Harlikwin]

Ok folks, This is the third iteration of the military model thread. The former/eaten mil-model II thread resides here if you would like to read it http://clash.apolyton.net/MilitaryThread1.shtml

Enjoy!!

-Harli

[Mark_Everson]

A Proposal for What Happens on the Boundary between Economic and Military Timescales

Well, we have had extensive discussions on how a "realistic" whole-history game can work. I'm not going to go over all the arguments that have led to our (me, Harli, korn469, and I think Krenske) tentative conclusion. If you want to see those you can check in a variety of recent threads including "Clash Scales", "Mil Model discussion thread", and various references therein.

A very terse summary of where we are is...
1. Everything that involves combat or strategic troop movement should all keep to the same "military timescale". To do otherwise would virtually guarantee completely unrealistic results some of the time.
2. The military timescale should be fixed over all time at one month of what could "really" happen. This means that at combat movement rates task forces will generally not be able to go more than a few squares (a few hundred kilometers). This is important so that the player can change grand strategy, and some operational factors, as a result of new information on how movement is evolving, and how battles fought have been resolved.
3. After locking in the decisions 1 & 2 it is Impractical to run the whole of history at this scale. And yet we want the economic results to be crudely realistic also. This mandates that at least for much of the game the economic timescale and the military timescale for a turn will not match. For instance, in the ancient world we might have the result that the economic timescale is one turn = 5 years.
4. For a game that covers all of history, we expect that the economic timescale will change drastically as history progresses. We hope to achieve parity between economic and military timescales for the final century or so of the game. Whether this is practical needs to be determined in playtesting later.
5. The major weakness of this system is in the areas of overlap between the models. For instance, picture that we are using five-year economic turns. This is the time period over which one might expect decent city walls could be made. However, since military task forces are only moving during one turn at a movement rate characteristic of a month, we could have the absurd case of defenders flash-building city walls as the enemy's army rides into view. The same could be true of a defender building huge amounts of arms and armor. In order to avoid destabilizing the game, and generating clearly unhistorical results, we simply need to make some adjustments at the joints between the models.

Note: to some of you this series of decisions may seem rather strange. I think that actually it is essentially what is typically done in Civilization right now. The approach was just never stated explicitly. That is why movement rates are so absurdly slow for the amount of time that is supposed to be passing in Civ.

In the remainder of this post I am going to address what I think is a reasonable approach for handling the joints between the models. I am specifically concerned about the interconnections between the economy and military models. As far as diplomacy and technology, I don't think we have any big issues. For characters, we will just have to see what "feels" best. Opinions on my ramblings below are actively solicited .

First of all, I've come up with three categories for how to handle the between-model issues. I denote them as...

M Follows the military timescale, FE resupply of combat troops with ammunition should only cost what a month's worth of ammunition costs, since you are only getting a month's worth of combat effectiveness out of it. There may be better examples, I just can't think of them.

E Follows the economic timescale, FE I think feeding and paying the troops should use up food and money at the same rate as people in the regular economy. If the troops only consumed one month of food and pay per turn, it would be very easy to maintain unrealistically large standing armies.

G The geometric mean between between military and economic timescales. I think something like this is needed for those cases where using either M or E would result in a big "huh?" from the player. For instance, let's say we are in the ancient world with economic turns that are five years long. So military turns are, as always, one month long, and economic turns are 60 months long. Let's use the example of building military hardware, say to outfit a legion. To produce these things will simply require getting a certain amount of production points built. If we adjust the cost to match the economic timescale, this hardware would come out way too quickly. However, if we adjust the cost to match the military timescale, that legion might require Forever to be built. A crude example follows...
Let's say we are in a small Roman province that produces enough taxes per year to just barely outfit a legion in the "real world". (taxes equal 100 C, cost of hardware for a legion equals 100 C, C equals the unit of money) And say our economic turns are five years long at this point in the game. So if we use "realistic" production rates geared to the economic system, this province could produce 500 C over the turn, and so pop out the hardware for five legions in a single economic turn. As noted above, flash building a sizable army will destabilize the military system, and we have to avoid that. So what if we boost the price to match the military system timescale? We would need to multiply the price by a factor of 60. This province would have to concentrate solely for 12 turns to produce the hardware for one lousy legion! So clearly using neither the military nor economic timescales to figure the cost is appropriate, we need something in between. That's where the geometric mean comes in. If we call building hardware G, the timescale we would use is approximately eight (~60 ^ 0.5) times longer than the military timescale, and eight times shorter than the economic one. So how does this work out? In eight months using the "real" numbers this province could build two-thirds of a legion's hardware. This seems to me like a reasonable balance that prevents obscene flash-building of armies, and yet doesn't completely cripple the economy to build a few units. I'm not sure the geometric average is exactly the right thing to use, but I think it's a reasonable thing to start with.

The things that I have come up with that fall between the two models, and the categories I am recommending, are:
E Regular Supplies: food, replacement parts, fuel, etc. that is not directly related to combat
M Combat Supplies: those supplies used to replenish/repair consequences directly resulting from battle (munitions replacement, medical supplies, etc.)
G Training: especially training to bring new recruits "up to speed"
G Arms, Armor & Munitions production
G Building Fortifications
E Economic Productivity of demobilized or "National Guard" type units

Well, that's what I've come up with so far! Please let me know what you think.

[Krenske]

Mark,
I agree that a standard should be set for a military turn. If only to have some sensible movement rates. This is most needed in the modern period.

How it integrates with a economic turn is indeed difficult to rationalise. For most ancient civilisations there was indeed a buy it when you need it mentality. If Rome needed a 50000 man army then it quite simply grabbed 50 000 men and gave them weapons. There was of course permanent fixed units but these were able to be quickly supplemented by rapidly formed armies. These armies could be formed in several ways in different civs. Materially the actual weapons of war could be stockpiled. (see Gaius marius's use of 60000 full sets of arms and armour to equip his plebian/free man levies, all of which was just sitting in warehouses.)

The difficulty is at the training and manpower levels. Any Civilisation has a military policy regarding military training. It may be neccessary to institute a base training level for the civ as a important SE factor. (All citizens of athens were required to maintain there own militia weaponry for example and had to train for a period while a teen. Similar occurred with Rome where tax/voting concessions were given for training.) Possibly a civ will have a base civilian training level of basic or trained instead of raw. Either way though the troops can still be called up and units formed at a rapid rate.

The major problem is one of available manpower. In any one turn there should be a basic available manpower value for the civ. To create a large army rapidly you must suffer in some way. Normally it is through a decrease in production or tax.

Note the following model is not really military but has military implications. I am sure similar models have already been proposed elsewhere so I will keep it simple.

This could be modeled with the following system. The available manpower value is the count of all those men (and later women) who are capable of being available for public duties whether they be in production or land improvements or wonder creation or even for use in the army. This available manpower should be broken into the four areas mentioned. Any not so assigned will earn tax wealth for the civ as they earn a living for themselves. There should be happiness penalties for assigning too large a proportion of this manpower to any one area. (Certain SE choices could decrease this happiness penalty. Assyria was very militaristic so it suffers a lesser penalty from military service.) Obviously other penalties come from decreased manpower in the remaining areas. Less in production gives less production, less free for tax earning gives less tax.

The proportion of the civs population in the available manpower pool should be based on civ advances and on SE selections. (With each increase in agricultural production more men join the available pool from the rural agricultural sector. Certain government types would be restricted in the amount of change they could make at once, etc.)


Back to the topic now. This would indeed allow for rapid army growth. Some basic examples of rapid army growth are.
Roman republic legion mobilisation in times of barbarian invasion. Forming 10 legions in the space of 6 months on occasion.
Persia calling up 150 000+ additional troops twice in the space of a year to counter Alexander.
China's 100000 man "peasant" armies that were called up for campaigns almost at will during the combined empire periods.
The US armies rapid growth from nearly nothing for the US civil war, again for WW1 and for WW2.
The British field army of WW1 was rebuilt after the loss of much of the BEF.

It is fairly obvious that when an army is needed by the civ it can be found and it can generally be found quickly the problem is what sort of pain the civ is willing to suffer, to do so. In some situations though there is enough manpower just "lying around" to do it ithout much pain. Rome managed to do it some times with the Dole recipients alone. US selective selection during WW2 allowed productivity etc to remain high as mainly the "non-productive" elements of society were pulled into the cutting edge of the army.

As a different point

To stop people sending a unit off into the unknown and using its enhanced movement to discover the world, Can I suggest the following. A rebellion/unhappiness feature based on the distance a unit is from a home city. This would restrict the distance a unit can move from the civ but not within the civ. This could be integrated to the various supply net ideas.

For exploration the civ could pay for explorers of 2 types. General explorers and specific explorers. General exploration just uncovers several squares of fog each turn. Specific exploration costs more but allows the exploration to be targetted. This would decrease micromanagement as each created explorer could just be assigned to one of these types of exploration and will just keep working until re-allocated. We would no longer need to move explorers and exploring units all about the place.

[Krenske]

As an addition to the above maybe, the initial call up/ creation of armies can be tied to the announcement of war. Like this, First a diplomatic message stating that your civ is at war is displayed and it gives you the option of going to a military mobilisation and creation screen. At this point any units can be purchased and changes to the civ manpower made. It is possible that the equipment is not available yet but at last the men start training. Then for the first year, units form rapidly as equipment becomes available. (Of course if equipment is stockpiled then the units form very rapidly.)

[Mark_Everson]

Krenske:

Agree with your basic points, but IMO they are already covered. I admit they aren't stated anywhere explicitly...

I already envision (and I'm pretty sure Harli does too) stockpiling of weapons going into units, and pre-training of the army or populace. Those units should only take 1 turn to form. (Probably defend immediately, but take one turn to be fully active)

As for your SE choices they should already be covered by the current tax system. In it the govt would tax people, and then buy training either for existing mil units, or "all available for service" so I think we can cover at least most of your points within the existing system.

The main thing I meant in my big post above is that you should not be able to take untrained civilians, and build weaponry and train them to a high level of combat effectiveness all in one turn, Even if the econ turn is 5 years.

[don Don]

I finally have time to spend going over the Clash combat models. Now that my butt and eyeballs have recovered here are some intial observations.

quote: Some elements may have other "special" parameters as well. • Number of turns away from base allowed: for AW elements… • Special move ability: Distance attack [artillery, naval bombardment], Amphibious, Air Drop, Underwater, Carrier, etc.

AAAARGHH! No, anything but "aircraft hanging in mid-air!" Please, please, have air unit attacks conducted as "Distance attack" with suitable range. I put mucho work into the movement thread for the Civ3 List (see Movement Summary). The ideas for movement and supply are general enough they can apply in the Clash system (your turn to wear out your eyeballs).

quote: ZOD: [Zone of Detection] defined in radius km, and depends on the size and tech level of the TF. … Friendly TF [FTF] may have overlapping ZOD, but the center of the a TF [TF Loc] may not be in another unit's ZOD. In other words, the TFs can be very close but not stacked.

[Diodorus Silicius mode]
This could well be a problem at almost any tech level. For a strategic scale military model with 1 year turns to work ground unit ZOD must be large, else the player is cursed with needing multiple TFs just to patrol one tile. But if ZOD is on the order of ½ tile width radius, then putting more than two TFs of any size in a tile becomes difficult (rather silly considering tiles are to be 100 km). For radio equiped units, air units, or naval units equiped with surface radar, air recon or combat aircraft the ZOD must be very large.

Air TF ZOD should be conducted as firm out to a tech dependent search range (easy flight distance for fighter types) and increasingly inconsistant out to "n" times the firm ZOD, "n" being tech dependent (max range special recon types) and reliability depending on how many search sorties the TF can deploy and what kind of units are present to be detected (e.g., battle of Midway). This could be true for ground and pre-flight naval TFs with as well, but that may be more nit-picky than desired.

In any case, the numerical size of the TF is already a given parameter. The physical sizes of TFs are more dependent on formation than anything else. One can pack 100k men into a stadium 500m across, house them in a few km² of city blocks, or deploy them over many km of the battle front. Dispersion should be based on doctrine/tech/unit type in a quantifiable way, thus you should know the TF physical size. Lesser units' sizes are negligible on the strategic scale, whatever their ZOD may be (e.g., a lone destroyer out on radar/sonar patrol, the seal squad on recon probe). Unit centers within ZOD should not be a problem unless they're trying to move one through the physical formation of the other. To some extent that can be handled by suggestions that follow.

Another matter: that corps deployed on the battle front no longer has a circular physical size or ZOD. Or move that same corps along a road, it will be two to four columns wide and miles long with little in the way of outriders/scouts on the flanks. The problem applies to a smaller ancient military unit stationed on your border sending out regular patrols to either side but not across the border or to the rear. The physical orientation is much more significant than the generic circular ZOD allows.

quote: • Front / Side / Rear: The TF faces the last direction of movement (or turn). This is used to modify combat parameters. … • Each fortification level yeilds a 5% higher Defense factor, and takes 1 ENG unit 1 turn to build. Up to 10 fortification levels may be built in a given 100x100 km sector. If there are no ENG elements, each fortification level takes 1 turn for the AG to build, but only levels 1-5 may be built without ENG units. • Fortifications persist after departure of the AG, and become terrain enhancements.

I'm wondering how much facing needs to be included on the strategic scale. When a unit is moving its facing may be germaine, but except for an initial contact round the game turns (and even movement segments) are too long to dictate that facing has a great effect on strategic combat. Typically facing is a purely tactical matter, of importance only after the units engage and can't maneuver effectively under fire.

Envelopment, flanking, mid-line penetration etc can take place on the strategic (or "operational") level that we can see on the map (e.g., Battle of the Bulge), but that can be folded into the supply model. Another case that must be taken into account at the strategic level is the army relieving the beseiged fort or city. Facing then becomes a strategic matter for the seige force trapped between the defenders and the relief force. The seige force may elect to move out to meet the relief force, in which case the city has been relieved (however briefly). That is, in essence, the primary purpose of a fortress or city walls: Delay the attackers until help from outside arrives or until they give up.

On the other hand, fortifications must have a specific facing, and only apply their benefits to defense from attacks in that direction. If you will recall the Maginot Line fiasco… of course, that's what the French get for leaving their defense to the AI .

I would propose that individual fortifications should not be abstracted as affecting the entire tile, unless you've deliberately built one big enough to do so. Each fortification built should have a specific x,y location just like a TF. Each fortification should have a capacity in terms of the number of units it can accomodate. Units at the same location as the Fortress would need a flag indicating whether they are "in" the fortification or merely encamped there. This could require dividing a TF between the Fortress and the redoubts or temporary defenses without.

Construction would be divided into "work units" instread of simple defense factors. Military units would have 1 WU output, labor or engineering units 2 WU output. Initial construction would be size 1 and level 1 on all faces with capacity for as many units as participate in construction, up to some limit (lots of "free" WUs). Such construction could be instantaneous under game conditions (Roman legions on the move in Gaul built an encompassing ditch/rampart/pallisade every night and raised it in the morning; I dimly recall a Civ1 patch that had instantly fortified Legions). Subsequent work either expands the capacity by 1 (level 1 defense, 2 WU cost) or the defense of one face (size*WU). The limit of level 5 would apply to Military or unskilled labor units, and such earthworks would only persist for a short time after abandonment.

The coordinate scale will have some effect. If set at 1 km then at some point additional capacity at one facing must become in essence a fortification at a new coordinate (linked in the database). Some size limit should apply above which a maintanence cost is required (continual expense of labor from occupying units, at the least). Permanent (stone, brick, concrete, massive earthwork) defenses require engineering units, some initial expense, but no maintenance while occupied. Perhaps truly impressive fortifications (defense >10 and/or huge capacity) could be built at suitable initial expense and maintenance cost (e.g., Caernarvon, Crak des Chevaliers, Maginot Line, Siegfried Line, Norad).

Building a Great Wall would actually be possible. (The GW of China wasn't manned, it simply served as a road and as a deterrent against the raiding barbarians taking livestock or captives.) City defenses should be treated similarly rather than abstracted (essentially fortresses of huge capacity). That would help to cure the "one size fits all city wall improvement" criticism leveled against Civ.

This is where the physical size/orientation problem could be resolved in many cases. By defining the temp/permanent defensive works you define where the units can be or can move to. A TF is either going to be on the line, behind the line in reserve (positioned roughly parallel to the line), or entirely out of the battle front area (physical position dependent on other factors). Moving through the defensive works in a passive confrontation or active battle situation is nigh on impossible.

I know I'm going over old territory again, but this could solve one of the headaches without too much additional coding (the same coding for moving/orienting units would handle the building, destruction, or delapidation of fortifications).

Another point raised (I won't go back to find a quote) is that of garrison forces. Historically, garrison forces were always small because 1) they could depend on the fixed defenses to multiply their effectiveness greatly, and 2) they had to be small for eco/social reasons (very few standing armies of any size in history).

Caerphilly castle could be succesfully defended with only 60 men against any army not willing to take horrific losses. More men would not help defense much; a factor of ten increase in manpower might only double the defensive strength, another factor of ten would not redouble defense strength. Once the gates are breached and melee commences in the wards then the number of defenders makes a big difference.

Between the French and Indian War and the American Revolution the Northwest Territories (Great Lake states) were held by a string of forts along Lake Erie and along the Ohio River. They were anchored at the northern corner by Fort Detroit, the largest with a garrison of 35 (count 'em, thirty-five) musketeers. Up through the F-P War Luxembourg was the key to controlling Europe. How big do you think the Luxembourg Army was? I believe it was DS who mentioned that in pre-Napoleonic Europe most armies operated within wagon-supply distance (150-200 mi) of Fortress depots. During and after Napoleon's campaigns Luxembourg remained in that role after most other fortresses fell or became obsolete.

Assaulting the fortified depot would only serve to bring the army in to your rear, so you had to detach a force to besiege the fort and meet the army in the field. Luxembourg was so centrally located and heavily fortified that it could swing a victory for whomever it chose to ally itself with. This only changed after the F-P War because everybody decided Luxemboug was too powerful and signed a treaty that called for the destruction of the fort!

After WWI the Belgians decided that getting rid of Luxembourg Fortress had been a bad idea, so they made their own version: Eben Emael. (Much better than that AI fortress jobbie built by the French ;¬D.) It might have worked too, except that the Germans managed to land a glider full of commandos on the nearly undefended top, a real stroke of luck. Also during WWII there was an American Division in retreat through a mountain pass that was covered by three volunteers in a well-placed bunker. These brave men delayed the pursuit by a day until a handful of German soldiers scaled the mountain and dropped in on their blind side.

We can't model the tactical aspects of fortresses but we need to be able to account for such effects on the strategic scale. The 5% x 10 levels maximum is insufficient without accounting for the essentially tactical bottleneck effects, too. The defensive benefits need to multiply the effective size of very small units otherwise insignificant on the strategic scale, but not "standard" strategic units.
[/Diodorus Silicius mode]

Timescale problems that don't seem to be resolved [hint, hint] but Mark wants to finalize it as is: Good strategic movement and production generalization helps. I have lots to say, but that's enough for now.
[This message has been edited by don Don (edited February 03, 2000).]

[Mark_Everson]

don Don:

Hi, glad to see you around again. Unfortunately you read the old model which is basically toast. Don't worry there will be no airplanes hanging in the air. (That was one of the first issues Harli, the new military model lead, raised when he came on board.) If you want to Really wear out your eyeballs you can read the Mil II thread on the web site (it died so a copy of it is there off the mil page). That is much more up to date. Hopefully in a week or so there will be a version of the new model...

I would really like to hear your opinion on the stuff in the Mil II thread and the Mil discussion thread (back about 2 weeks). But best wait a little bit for your eyeballs to recover.

Cute on the DS mode

[Krenske]

Don Don,
Don't worry about planes sitting in mid-air, I am fairly certain that a system based on influence at range will be used not unlike your recon example.

Air units will probably operate on very short cycles. My proposal that I am formulating has the basic air unit as a wing (4+ squadrons). The wing will be allocated to a role on a yearly basis and will conduct operations based on that role for the period. (A bomber unit could be allocated to maritime interdiction. So for the next year it attacks shipping and conducts recon over oceans within its area of influence.) The unit can also be given a tempo of operation, either continuous or wave. Continuous means the unit attempts to carry out its role every day of the year (actually each basic combat turn, I know it won't be days). Wave means the unit will only carry out a mission when most of its strength is available. (The classic example of this is the bombing cycles/surges of british bomber command during WW2 where the heavy wings would surge for 2 weeks and then standown for maintenance for 2 weeks.)

The continuous tempo units will maintain a presence at a reduced strength the wave tempo units will appear enmass but only intermitently.

There will be more coming, as soon as harli and I hammer together the basic ground combat model.

[LOGO]

First of all, Mark sorry I havn't posted lately but I've been very busy. Although I don't think I've really missed much (posting has seemed to slow down lately)

Anyway, I'm confused when you say we can't have a 1M to 1E turn in the game? When you say it's impractical do you mean in the AI?I've tried to look over older posts but I still don't quite get it.

Second, do we have a definate land combat model yet? Everyone has their own ideas (including me) about all the variables, I think the most important thing we can do right now is hammer out the most simple of all systems for this game, also remember that one of the huge complaints about Civ. II was the totally unrealistic-odviously not fully thought through-combat system. So come on guys let's get that.

Third (misc.), I really think there should be a basic limit to how quickly you can produce human units because of training-development of things like VR systems would increase that.

There should also be a drafting system, for example in some societies you have to wait for people to freely sign up for the military e.g. America.

You should need to create a reason to fight a war (espiacally in a democrasy) popular choices will include holy crusades, they attacked us first, keeping "sability", liberating the poeple. And you should be able to stop them from making a reason, for example Germany never invaded Turkey in WWII because of their shrewed diplomacy never gave Adolf a reason too. This should effect how allies help the war effort.

[Mark_Everson]

Hey Krenske:

Pleeeze don't say year for military turns... it is Always a month worth of action. (unless the concept turns out not to work) Say turn or perhaps month!

Logo:

I really don't know what more to say on the issue of 1M = 1E turn. I have written Pages on the genreral area. One final bit...

It was reasoned in conjunction with the military guys and our proposed movement system that 1 operational month is about right for a 'Military turn'. This guarantees generally sensible things should happen in the mil area. If 1 month = 1 Military turn = 1 Economic turn then if starting at 5000 BC we have 7000x12 = Nearly 100,000 turns!!!!!!!!!!!
Doesn't that sound like a bit of overkill? I would like to be able to play Clash through more than once in my lifetime.

If you want to drop the 1 mil turn = 1 month then you can say your piece, but it is basically locked in. So unless there is a Very persuasive demonstration that it can't work we are going with it.

Your point on war rationale discussion belongs in govt...

[don Don]

Yes,. Mark, I did read the Mil II thread & Mil discussion in their entirety, thus the need for butt/eyeball recouperation. I didn't see anything specific about the air model, only compaints that someone was going to work on it but hadn't blah blah. So I thought I'd make doubly sure it was addressed!

I said I had lots to say, and I do

I want to resurrect something from the old Military/Combat thread (with modifications from intervening comments, etc):
----------
One way of incorporating doctrine, training, and historical factors is through distinct organizational tech advances. Just to throw the idea out for discussion, stripped of any numerical bonus/penalty specifics:

Organization Level   Unit Size   Description
ALL ERAS
·Band                    1       Loosely organized unit
                                 (std primative production unit)
·Horde                   4+      Large (unwieldy) group
· Stub                   ¼       Disrupted unit fragment
· Mob                    ¼       Spontaneous disorganized unit
· Partisans              ¼       Spontaneous paramilitary unit
· Militia                ¼       Part-time military unit
ANCIENT ERA
·Troop                  1-8      Standard ancient production unit
·Phalanx                 4       Aggregate foot soldier unit
  Company                ½       Phalanx subunit
·Myriad                  8+      Aggregate foot or cavalry unit
·Legion                 4-6      Composite foot & cavalry unit
  Maniple                1       Legion composite subunit
  Century                ¼       Legion fractional subunit
MEDIEVAL/RENAISSANCE ERA
·Troop                  1-4      Standard med/ren production unit
·Battle                 4-8      Aggregate foot or cavalry unit
MODERN ERA
·Battalion               1+      Standard modern production unit
  Company                ¼       Battalion fractional unit
·Regiment               2-4      Aggregate unit
·Brigade                4-8      Aggregate unit
·Division               16+      Aggregate unit
·Corps                  32+      Aggregate unit


The general idea here is that each ancient, medieval and modern organization type is a separate advance involving C³ and supply issues that we can generalize this way. Weapons are a somewhat separate issue, tech-wise.

At first all military units are Bands. If you ever get 4 or more Bands together in the same place doing more or less the same thing they become a Horde (they can't help it). A Band would suffer some minor A/D penalty because of lack of discipline, communication, etc.; a Horde would suffer a larger penalty.

A Stub (or a collection thereof) is what's left over after a military unit is defeated. They can be regrouped to form replacement units or used to diminish the cost of new production units. Mobs and Partisans should be self-explanatory. These fractional-sized units suffer A/D penalties based on size and morale status (likely to be low, except Partisans).

The Troop advance enables what we would consider very basic military discipline and training. They can be grouped together without suffering command problems (no penalties). Then the Phalanx represents tactical training, which confers a minor defensive bonus. The Myriad and Company (perhaps there was a different term?) represent further advances in training (think Alexander the Great here) with A/D bonus. After a battle, broken unit stubs could be reorganized into whole units with some morale penalty. After development of Company no morale penalty would apply to the reorganized unit because of stronger NCO structure.

The Legion advance covers the primary tactical and discipline level of the Roman Legion (but suited for whatever culture is relevant, just use a different name). This was the first true "combined arms" unit and would have a substantial A/D bonus. Legion Subunits are further advances enabling flexible use and deployment, Maniples having some minor bonus over the basic Troop, and Centuries having lesser penalties than Partisans or Stubs and organization benefits as described for the Hellenic Company.

Therefore a Legion of (2 hv inf, 1 lt inf, 1 lt/md cav) would be better than a TF of 2 hv inf Troops plus 1 lt inf Troop plus 1 ly/md cav Troop. After development of the Century you could attach fractional support units (a hv cav or an elite infantry) for additional bonuses. Players could while away hours fidgeting with unit composition or settle for AI defaults. The Roman Legion was possible in part because the patrician landowners who constituted the Legions of the Republic considered the footsoldier to be an honored role. In Japan a similar structure existed, the Samurai were primarily unmounted knights, but this did not result in such an ingenious combined arms unit. In some ways I'd like to see this reflected, as well as the social structure that resulted in the Mongolian Tumans.

My Medieval/Renaissance model is necessarily sketchy, as I am less familiar with strategic level stuff for that time period. (Where the heck is Diodorus Silicus when you need him?) I am even less familiar with military organizations of India and the Far East. In general it seems that the combined arms approach used by the Romans didn't pass to the Germanic tribes due to their class structure. Infantry was largely peasant and cavalry exclusively gentry, never the twain to meet. The Legion model with some improved weapons could simply continue up to the Musket (and perhpas beyond) should a civ not fall into the Germanic model.

Modern Era aggregates also represent levels of military doctrine and training with increasing A/D bonuses. With the development of Battalion tech, the aggregates would be available without penalties or bonuses. After developing Regiment, all larger aggregate units would receive only the Regiment A/D bonus, with Brigade all larger aggregates recieve only Brigade bonuses, etc. Modern mixed units (i.e., infantry support for armor) are handled at the TF level only, not as a distict organization as the Legion was in the Ancient era.

Construction of units would be flexible. If you wanted to produce a garrison force for a border outpost you could build a size 1 unit and place it there. Rural villages could produce fractional-sized Militia for defense. Small cities could produce fractional-sized subunits for garrisons once that organizational developement is discovered. You could produce a Troop of up to size 8 as though it were one unit. For a Phalanx you could either build one directly or take size 4 worth of Troops and train them into a Phalanx.

It would be possible to either tie the units directly to single population units (for Partisans/Militia, for example) or to a group of population units. If you build more than the population growth can replace in the time-span of a game turn, then producing a military unit will "consume" a population group of the same size. Also, disbanding a unit requires finding civilian employ for the artificial population "growth" that occurs. The effects would vary depending on government/social system.
----------
A primary concern in time scale between Mil & Econ turns is to limit the ability to build units faster than the military is permitted to move, and I think there are ways to handle that better than simply fixing the mil turn at 1 month and then applying a mathematical fudge factor to the production capacity to compensate. I bring this up here because I am dealing solely with the issue of military unit production, rather than production in general. That, and it seems that the Mil guys are the ones driving the Mil/Econ ratio issue.

It would be better if the production system inherently limited the ability to instantly summon real troops (levies are a different matter entirely), and the movement model allowed for strategic movement to be separated from combat/exploration etc. I think that the Mil Discussion got bogged down because nobody wanted to separate the "stepping" of the units through movement from the length of the military turn and the ability of the player to intervene either with orders or with production.

Perhaps it would be better to leave open the possibility of reconciling the time scales, or at least coming closer than 1 month Mil to X years Econ. There are so many possibilities. Believe me, I deliberately limited my own suggestions in the Civ3 List material on movement to those things which I thought had a passing chance of being implemented in a Civ2/SMAC style game. Nobody has really addressed the issue of generalization in the movement & supply (not support) model for Clash that I have seen. But that isn't the focus of this post.

I've mentioned before, as an old paper map and cardboard chit wargamer one of my favorite games is Global War (S&T ~1975). It is a truly elegant WWII simulation. The production is generalized to the whole country rather than to individual production centers. This is shared by the Clash model at the province level. How does GW limit production? Each turn the production of one type of unit (out of 2 ground, 3 air, 5 sea, and 3 special types) can only exceed production of that type in the previous turn by 1. Something along this line would work very well in Clash.

So, if I have a relatively undefended city and a TF lands a couple tiles away I can't rush 8 military units in one turn no matter what my capacity may be. I might be able to produce 1 foot and 1 cav unit at most (1 each of the two types). Then some allowance can be made for levy troops, dependent on government & social structure.

In the table above certain units can be produced in "bulk" as though a single production unit. For GW the units were Division size (although some were labled as Corps IIRC), rather large units, with no consideration for increments below that. In Clash the size would be limited by population size, general mobilization level, and gov & soc factors. So now a larger city in the above situation might be able to raise a size 2-4 foot unit and a size 1-2 cav unit to supplement a small garrison. In essence we do not use a rigid definition of what "1" unit is. [Yes, I've looked over Druid2's old stuff and Krenske's newer stuff, but i don't want to broach that now. Let's assume reconcilable models at this point.]

Civ simply allows for large standing armies that couldn't have existed in a vaguely realistic economy; Clash can allow for the quick production of limited defense closer to reality and tweaked for gameplay. A raiding TF might be thwarted at that level of defense, or a larger TF delayed until help arrives.

If the econ model works well it will not be like Civ where you can build up a huge treasury. I think historically that would be exceedingly rare. This will also serve as an effective limit to how much can be produced on the spot. So if my province can raise 100 in taxes, which is the cost of a full Legion, can I truly forego paying my civil employees and other expenses and throw it all into Legion production? Not in every circumstance. If my province has a total reserve and surplus of 100 I might still not be able to raise a whole Legion due to time constraints.

I will go to GW again for a bit of inspiration: the production of each type of unit takes a fixed number of turns. So in Clash units will not appear instantly on the Mil time scale. For large ships even Econ turns may be shorter than production time. Then we don't have to worry about production when reconciling/scaling Mil & Econ turns regarding movement.

[Harlikwin]

Hi all,

Before everone goes nuts about the combat model, Kerenske and I have worked one out for th most part and will be handing it over to mark to code. I will also get a copy put up on the web site. Most issues I have seen on the fourm are adressed somehow (even if it is a we don't care cuz of scale answer) Hopefully It will be done in the next week or so (Hopefully sooner)

-Harli

[Mark_Everson]

Harli:

Even a draft as early as its coherent posted here would be useful, so I could start thinking about the general approach to coding it. Also that would give people a chance to see and comment on the basic outlines, which could save you work if there's a problem in it.

don Don:

The Global War build limitations sound intriguing. I had been thinking about something like this anyway so that you couldn't immediately change 100% civilian production to 100% military.

Harli, Krenske, or don Don:

If you have any numbers on strategic (friendly territory) vs combat movement rates on different terrain with and without roads I'd like to see it. I may have to fake up something fast in this general area for the demo.

[Henrique Duarte]

"Don don
The Legion
This was the first true "combined arms" unit and would have a substantial A/D bonus"

IMO the Philip II Phalanx/Peltastoi/heavy cavalry army organization was the one to introduce and adopt well thought combined arms tactics. The problem with this organizatoin was flexibility - a phalangite could never fill the role of light or even regular heavy infantry in open order. If you lacked any of these components in the battlefield... your dead (Cynoscephalae - hope I spelled this one right:-))

The great inovation of the romans (specially with Marius reforms) was the tremendous flexibility of the legionare. He did everything from transport of equipment (Marius Mules) to skirmishing in light order to close combat or even tight formations to meet cavalry charges (not to mention their engeniering role - they built several of the roads of the empire!). As long as discipline and training was high the roman legionaire would assume the role most needed on the battlefield.
In fact such was the supremacy of the legionare that the romans "almost" ignored the cavalry component of it´s army...and they payed dearly for that in Cannae, against the parthians and in late imperial period when they needed large quantities of quality cavalry to meet the huns & friends.

Still the idea you sketched for Task Force organization limits according to tech is quite a neat one and I hope the mil fellows give it proper consideration.

By the way, do not forget that the romans had a highly motivated citizen army in early and mid republican times and a PROFESSIONAL army in late republic and imperial periods.

One last thing - I hope someone touches the Assyria problem - they had the first and only permanent professional army when everybody else was mobilizing/demobilizing militias. I believe the game model can come close to simulate such situations but it is better to write this somewhere so that it isn´t overlooked.

[Henrique Duarte]

another thought

"Don don
So, if I have a relatively undefended city and a TF lands a couple tiles away I can't rush 8 military units in one turn no matter what my capacity may be. I might be able to produce 1 foot and 1 cav unit at most (1 each of the two types"

I think this could work:

Military management
The player defines its ideal standing army order of battle.
ex: create 2 task forces with ideal composition of 4HI 2LI 1LC each and place them on pronvinces X Y
The moment the player creates this TF the AI will look at available Manpower (in this case it must be trained recruits) and weapon stockpiles and add as much as it can to those task forces. If you have enough weapon stockpiles and trained manpower you may be able to field huge armies nearly instantly (the romans sometimes raised a consular army every year 20.000 men, while in Yom kippur war the israeli mobilized in a few days some 300.000 men). If you don´t have either the manpower or the equipment, the output of your military infrastructures and production facilities (ex. barracks and factories) will determine the replacement rates in which you forces will be replenished.
This management would be rather simple for the player and would allow for much more realistic game situations. Raising an army was never that dificult (taking game turns after game turns to raise a warriors unit in CIV was a complete turn down) maintaining, handling disatisfaction and facing the economic repercussions of that army...now that is the big problem.